Greetings all,
I have a question regarding running and timing some machine learning models. I would like to do (what my buddy and I call) a "bake off".
I would like to measure time intervals for the different ML models using the same dataset using either Sys.time()
, system.time
or library(rbenchmark)
. Is one method preferable?
Any other pointers would be interesting.
P.S. one more item; I am using a set.seed()
for the testing. It seems to me that is something that I would want to Keep for this type of testing. Is that a correct assumption?